Nvidia to Invest Up to $100 Billion in OpenAI as Data Center Arms Race Intensifies

In a landmark move underlining the accelerating arms race in artificial intelligence infrastructure, Nvidia has announced plans to invest up to $100 billion in OpenAI. The deal involves OpenAI purchasing Nvidia’s advanced AI chips, and in return, Nvidia will acquire a non-controlling stake in OpenAI. The funds are earmarked for massive scaling of data center capacity – building facilities able to deliver at least 10 gigawatts of compute power, a scale comparable to multiple nuclear reactors. The first tranche (about $10 billion) will kick in following deployment of the first gigawatt of new facilities, and the companies aim to begin delivering systems by late 2026, leveraging Nvidia’s upcoming “Vera Rubin” chip architecture. This expansion is being driven by surging demand from frontier model training, large-scale generative AI services, and deployment needs globally. While the deal solidifies Nvidia’s position in the AI hardware supply chain, it also draws regulatory and strategic attention given its potential to reshape competitive dynamics among hyperscalers, cloud providers, and AI research labs.